Environmental Engineering Reference
In-Depth Information
Table 1.1
(continued)
Sampling
Selection of a subset of potential experimental units from the target
population for measurement of variables of interest
Sampling error
The variation among samples (or observations) of a given experimental unit
Sampled
population
Population from which samples are taken
Statistical
inference
The process of drawing sound and appropriate conclusions from data subject
to random variation
Target population
Population for which inference can be made
Theory
A broad, general conjecture about a process that can be tested using study
design
Treatment
Something that an investigator imposes on experimental units in some
deliberate manner
Unbiased
Long-term average of sample estimates equals population value
A principal goal of study design is to minimize personal bias, values, beliefs, and
subjectivity of the scientist so that conclusions can be supported beyond a reason-
able doubt. Basic tenets of modern study design are grounded in statistical theory
and have been applied for
>
75 years (Fisher
1935
). Exponential increases in
computational ability during the past 25 years have allowed for increasingly
complex approaches to study design and data analyses. However, failure to adhere
to basic components of study design cannot be overcome even with the most
complex analytical tools. A well-designed study will allow investigators to focus
on current knowledge gaps, provide rigorous tests of information, and enhance
efficient use of resources.
1.2 Role of Study Design in the Scientific Method
Study design is a critical part of the scientific method (Fig.
1.1
). There are several
variations of the scientific method for studying ecological systems, but all include
steps of (1) construction of question(s) that address uncertainties in the ecosystem
of interest, (2) formation of theories to explain observations or questions based on
observation, which leads to multiple hypotheses that have predictions suitable for
testing data, (3) design of a study to test primary and alternative hypotheses and
their associated predictions, (4) collection and analyses of data, (5) report
conclusions and make inference from results, and (6) communicate results through
the peer-review process that adds credibility to the findings (Gauch
2003
). Typi-
cally, the scientific method is referred to as a process because all studies and
resultant conclusions are subject to being repeated, typically as a feed-back loop
restarting with step (2) by other scientists. However, it would be a mistake for
anyone to perceive the scientific method as a defined sequence of steps leading to
knowledge, but rather as a framework for creative and productive processes that can
be used to accumulate knowledge that leads to truths corresponding with reality of
